Count the exits
Open your landing page and count everything clickable: nav items, footer links, the logo, social icons, the secondary button, the “learn more”. That number is how many ways there are to not do the thing the page exists for.
This is not an argument for removing navigation from every page, a site people browse needs it. It is an argument for knowing the number, because on most landing pages nobody has ever counted.
The useful reframe: a page is not a place, it is a step. Steps have one exit forward. Everything else on the page is either supporting that step or competing with it, and there is no third category.
Once you see it that way, the cuts make themselves. The footer link to careers is not evil; it is just not helping this particular visitor do this particular thing.
The single-CTA page
Zero sells air conditioning: a product nobody photographs well and nobody gets excited about. The page is one screen: a room you would want to be in, the unit sitting quietly in it, and a gradient sweeping the wall that is the airflow drawn as light rather than as arrows.
One promise: feel the difference. One button, designed in both rest and hover, because the button is the page's only job and it deserves the same attention as the hero.
Everything the page could have said about compressors and efficiency ratings is absent, not because it is unimportant, but because none of it is what makes someone press the button.
When one page is wrong
When the visitor needs to compare. Pricing tiers, plan differences, model ranges, comparison needs room, and cramming it into one screen produces a page that is technically single-purpose and practically unusable.
When there are genuinely different audiences with different next actions. Forcing them through one funnel means serving all of them badly, which is worse than a menu.
When the purchase is slow and researched. Some buyers need six pages before they will send a message, and refusing to write them does not speed anyone up.